The Exemption That Isn't

The email from the Arizona Health Care Cost Containment System — AHCCCS, the state’s Medicaid program — will arrive sometime between June 30 and August 31. It will explain a new requirement. Starting January 1, 2027, adults on Medicaid expansion must document at least 80 hours per month of work, school, job training, or community service to keep their coverage. There are exemptions. One of them is for substance use disorder.

But the exemption is not what it appears to be.

On June 1, 2026, the Centers for Medicare and Medicaid Services issued an interim final rule implementing Medicaid community engagement requirements, the work requirement provision passed as part of H.R. 1. The rule lands during the most consequential moment in addiction treatment coverage in a generation. According to an analysis by the Center for American Progress, more than 1.6 million people currently receiving Medicaid-covered SUD treatment will become uninsured under these requirements — even accounting for the carve-outs. Another 1.72 million who aren’t in treatment now but will one day need it won’t be covered when they get there.

These are not theoretical projections. They are people. And the gap between what the rule says and what it will actually do to them is the story that behavioral health providers, policymakers, and anyone who cares about the overdose crisis needs to understand right now.

The Exemption That Narrows Itself

The rule’s text reads like protection. People who are “medically frail” are exempt from community engagement requirements. And the rule defines active substance use disorder treatment — participation in a drug or alcohol rehab or recovery program — as qualifying someone for that medical frailty designation. On the surface, it looks like CMS threaded the needle: you can’t be penalized for an illness that prevents you from working.

But that’s not quite how the rule works. CMS explicitly interprets the medical frailty exemption to require not only the presence of a qualifying condition, but also that the condition significantly impairs the individual’s ability to satisfy the 80-hour monthly requirement. That’s a functional limitation standard, not a diagnostic one. A person can have a documented opioid use disorder and still fail to qualify for the exemption if a state Medicaid agency determines their condition doesn’t rise to the level of functional impairment the rule requires.

This is what the Center on Budget and Policy Priorities called the “last-minute restriction” that will worsen the rule’s impact: CMS narrowed the exemption language in its final rule compared to earlier versions, tying the protection to function rather than diagnosis. “It adds a layer of evaluation that didn’t exist before,” the CBPP analysis notes, “and that evaluation will be conducted by state agencies operating under budget pressure to reduce enrollment.”

There’s a second trap door. The rule contains a provision that individuals who have been in recovery for five or more years — five years sober, stable, functional — may no longer qualify for the medical frailty exemption at all and could be subject to “individualized review.” The rule’s logic: if you’ve maintained recovery for five years, maybe you’re not impaired anymore. The practical reality: you’ve maintained recovery because your coverage has kept you in treatment and stable. Disrupt that coverage, and the recovery timeline that earned you out of the exemption suddenly becomes the thing that kills it.

Brookings Institution researchers have named this exact paradox. Writing on the structural problem with applying work requirements to SUD populations, they observe that the disorder itself impairs the ability to maintain stable employment — which creates a circular trap: the illness that should qualify you for exemption also prevents you from performing the activity you need to demonstrate to claim the exemption. Adding documentation requirements to that loop doesn’t simplify it. It turns bureaucratic complexity into a filter that drops people off coverage independent of their clinical need.

It turns bureaucratic complexity into a filter that drops people off coverage independent of their clinical need.

What the Documentation Burden Actually Does

CMS permits self-attestation through January 1, 2028 — meaning for the first year, individuals can declare their exemption status under penalty of perjury without providing supporting documentation. After that, states must verify exemption claims using “reliable information.”

That sounds manageable. But behavioral health providers know what documentation burden does to vulnerable populations. SAMHSA’s own data on treatment engagement consistently shows that the single biggest predictor of dropout from SUD treatment is administrative friction — copays, prior authorization delays, appointment complexity. The research on Kentucky’s Medicaid work requirement experiment in 2018, before it was struck down, found that roughly 95,000 people lost coverage within months, not because they didn’t qualify, but because they couldn’t navigate the reporting system. Many of those people had mental health or substance use disorders.

The same machinery will run again in 2027. States are required to implement by January 1, 2027, though they may implement sooner via 1115 waivers. And the current rule leaves enormous interpretive latitude to states — how they define “actively in treatment,” how frequently they require exemption documentation, whether they accept a letter from a treatment provider or require something more formal. KFF’s analysis of the rule notes that this variation “creates a coverage patchwork” where an identical clinical situation can result in maintained coverage in one state and disenrollment in the next.

Arizona is not starting from a position of strength. The state ranks 49th out of 51 for behavioral health access, and fewer than one in twenty Arizonans with opioid use disorder receive medications like buprenorphine or methadone. AHCCCS covers all three FDA-approved medications for OUD — a meaningful commitment — but that commitment is only worth something if enrollment is maintained. A person who loses Medicaid while in buprenorphine treatment doesn’t gradually step down care. They stop. And when they stop, the probability of overdose spikes sharply in the weeks that follow.

The People the Math Misses

The 1.6 million figure from the Center for American Progress — people currently in Medicaid-covered SUD treatment who would lose coverage — is the number most often cited. It is not the most important number.

The most important number is 1.72 million: the people who are not in treatment right now, but who will need it. The people who haven’t yet called a crisis line, submitted an intake form, or walked into a detox. The people the treatment system is built to catch when they finally reach out. Under the new rule, many of them won’t be covered when they do.

This is where the psychology of treatment-seeking matters more than the policy text. People with active substance use disorders don’t seek treatment on a schedule. They seek it in windows — acute moments of motivation, often following an overdose or a loss or a legal crisis, when the ambivalence that characterizes addiction briefly tilts toward help. Those windows are narrow. A week, sometimes days. Treatment research going back to the 1980s shows that the single strongest predictor of whether someone enters treatment when they seek it is whether a bed or a provider slot is immediately available. What the research doesn’t have to explain, because anyone who has worked in treatment already knows it, is what happens when the other barrier — insurance — closes while the window is open.

Denying coverage to people who are not yet in treatment, because they can’t document an exemption for a disorder they haven’t been formally treated for, is not an administrative technicality. It is the policy equivalent of locking the door.

What Behavioral Health Providers Need to Do Right Now

The rule’s comment period is open. The interim final rule is not fully settled — behavioral health organizations, provider associations, and patient advocates have the opportunity to submit formal comments before the rule is finalized in its permanent form.

Beyond advocacy, providers need to move on operationalization immediately. The behavioral health organizations flagged as most at risk by the Behavioral Health Business analysis are those without dedicated care coordinators who can help patients navigate the new documentation requirements. The pattern from Arkansas and Kentucky’s prior work requirement attempts is clear: without active outreach from clinical teams to help patients file exemption paperwork, the people who most need coverage are the least likely to keep it. The disease itself makes self-advocacy harder. That is not a character flaw. It is the pharmacology.

For providers with Medicaid-covered SUD caseloads, the action items are specific: document treatment participation clearly in records now, before the attestation requirement kicks in; identify which patients fall in the 5-years-plus recovery window who may face individualized review; and build workflows to assist patients with the June 30–August 31 outreach letters that AHCCCS and other state Medicaid agencies will begin sending.

For Arizona specifically, Chris and Pat on the Rize team should track the AHCCCS implementation timeline closely and flag it for the provider network — this is going to generate a surge of confused and frightened patients, and the facilities that have clear processes for coverage navigation will become essential referral destinations. The opioid settlement funding that Maricopa County is distributing should explicitly include navigation support for coverage transitions.

The opioid settlement funding that Maricopa County is distributing should explicitly include navigation support for coverage transitions.

The Rule Says It Protects People with SUD. That Is Not What It Does.

The medical frailty exemption in CMS’s Medicaid work requirement rule was designed, at least in its rhetoric, to protect the people most damaged by the substance use crisis. The implementation details — functional limitation standard, five-year recovery cutoff, documentation burden, state interpretive latitude — systematically undermine that protection for the exact people it was supposed to cover.

There is a term for a policy that looks like a safety net but has holes precisely where the most vulnerable people stand. It is not a safety net. It is a test that penalizes illness, and the people who fail it are not people who didn’t try hard enough. They are people who needed a functioning healthcare system to show up for them. On June 1, 2026, that system formalized a new set of conditions under which it won’t.

The letters start going out in 26 days.